Pith. sign in

REVIEW 3 major objections 4 minor 1 cited by

The Dipole Instability in Gravitational $N$-body Systems: A Natural Explanation for Lopsidedness and Off-Centered Nuclei in Galaxies

T0 review · 3 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read A sharp density break in a spherical galaxy destabilizes the halo into a rotating, sloshing $\ell=1$ mode that dislodges the central cusp and can explain lopsidedness and off-center nuclei.

desk verdict A well-supported linear instability with a saturation mechanism that is plausible but not fully proven; worth refereeing carefully. read the letter →

arxiv 2505.23905 v1 pith:EIRXXFLV submitted 2025-05-29 astro-ph.GA

Shashank Dattathri , Frank C. van den Bosch , Martin D. Weinberg , Uddipan Banik This is my paper
classification astro-ph.GA
keywords galaxydynamicsdipoleinstabilityl=1modedistributionfunctioninflectiondoublepower-lawdensityprofileN-bodysimulationssolitonlopsidedness
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper tries to establish that a common family of galaxy and dark-halo density profiles is dynamically unstable on its own, with no external trigger. When a double power-law profile transitions sharply from a steep outer slope to a shallow inner slope, the isotropic distribution function of orbits develops a 'bump,' a small energy band where ${\rm d}f/{\rm d}E > 0$, which violates Antonov's stability theorem. In high-resolution $N$-body simulations the bump turns the otherwise weakly damped $\ell=1$ dipole mode into a rapidly growing one: the mode rotates, torques the central cusp off center, and saturates into a long-lived soliton that keeps sloshing through the core. The same resonance physics drives the bump-on-tail instability in plasmas. If real systems pass through such profiles as they form cores or as their inner slopes are altered, this would explain persistent lopsidedness and off-center nuclei in galaxies.

What carries the argument

The object that carries the argument is the bump in the isotropic distribution function $f(E)$: an energy interval where ${\rm d}f/{\rm d}E > 0$, obtained by Eddington inversion of the $(\alpha,\beta,\gamma)$ density profile. In the linear response matrix $M(\omega)$, the resonance denominator $1/(\omega - \ell\cdot\Omega)$ picks out exactly this gradient; with a positive gradient at the corotation resonance the $\ell=1$ Landau mode is reinforced rather than damped. In the nonlinear regime the same resonance traps particles into librating orbits, and the trapped-particle dynamics, the soliton, is what erodes the bump and sets the saturation amplitude.

What would settle it

Build the $(\alpha,\beta,\gamma)=(6,5,0.5)$ spherical model with the same density profile but replace the bump by a flat plateau in $f(E)$, then evolve it with at least $10^6$ particles; if a growing $\ell=1$ mode still appears, the inflection is not the cause. Conversely, the paper predicts the unmodified model should show exponential $\ell=1$ growth with ${\rm Im}(\omega)=0.044$ and pattern speed ${\rm Re}(\omega)=0.846$ in model units, so a high-resolution run that shows no such mode would refute the claim.

Watch

Extended reading notes

Core claim

The central claim is that isotropic, spherical, self-gravitating systems with a double power-law density profile are unstable to a rotating dipole ($\ell=1$) mode whenever the transition between the outer and inner slopes is sharp enough to create an inflection in the Eddington-inverted distribution function. The paper shows that such an inflection causes the linear response matrix to develop a growing zero at complex frequency $(0.846, 0.042)$ in model units, in close agreement with the exponential growth measured in the simulation. The growing mode dislodges the central cusp and carries it outward on a spiral before the pair settles into a long-lived soliton that rotates and sloshes through the center on a slowly precessing, moderately eccentric orbit. Resonant particles trapped at corotation and at nearby Lindblad resonances librate and exchange energy with the mode, eroding the bump in the distribution function; saturation coincides with the near-erasure of the inflection.

Load-bearing premise

The whole analysis assumes the system remains near enough to a spherical, isotropic equilibrium that the distribution function can be recovered from the spherically averaged potential and the density of states; if the real system is strongly anisotropic or the mode amplitude grows beyond the few-percent level, the measured bump erosion and the inferred mechanism could no longer be trusted.

Editorial extensions

If this is right

  • A wide part of the $(\alpha,\beta,\gamma)$ parameter space is inherently unstable: profiles with small $\gamma$ and large $\alpha$ (shallow inner slope, sharp transition) violate Antonov's criterion, while the familiar cuspy profiles sit safely in the stable region.
  • Cored profiles lie close to the stability boundary, so any process that turns a cusp into a core can push a system into the unstable regime and trigger a dipole mode.
  • The resulting lopsidedness is long-lived: the saturated soliton keeps rotating and sloshing for the full duration of the simulation, so off-center structure need not be a transient disturbance.
  • The saturation amplitude of the mode grows with the depth of the distribution-function bump, so the observable strength of lopsidedness encodes how strongly the stability criterion is violated.
  • Because resolving the mode requires roughly $10^6$ particles, lower-resolution cosmological simulations are expected to miss this instability entirely.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Extension: I would predict that core formation by feedback-driven outflows, which lowers $\gamma$, can actively drive a halo across the stability boundary; the paper lists this as plausible but does not simulate it.
  • Extension: A direct test would be to look for coherent $\ell=1$ sloshing of the dark-matter density peak around the galaxy barycenter in cosmological zoom simulations with about $10^6$ or more particles per halo, since the paper shows the mode is invisible at lower resolution.
  • Extension: The same resonance balance implies that dynamical buoyancy of a massive perturber in a cored halo is the finite-amplitude counterpart of this instability; measuring whether the buoyant phase corresponds to a positive ${\rm d}f/{\rm d}E$ at the perturber's energy would link the two phenomena.
  • Extension: If off-center nuclei in dwarf galaxies are caused by this mode, their offsets should precess with a pattern speed tied to the central potential rather than wander randomly; tracking the phase of the offset over several dynamical times would distinguish the two cases.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 4 minor

Summary. This paper investigates the stability of isotropic, spherical systems with generalized (alpha, beta, gamma) double power-law density profiles. The authors show that for rapid inner-to-outer slope transitions (large alpha), the Eddington-inverted distribution function develops an inflection with df/dE > 0 in a region of energy space, violating Antonov's sufficient stability condition. They map the stable, unstable, and unphysical regions in (alpha, gamma) for fixed beta and provide fitting functions for the boundaries. Using multimass N-body simulations of the fiducial (6,5,0.5) profile, they find an exponentially growing l=1 (dipole) mode whose growth rate (0.044) and pattern speed (0.829) match linear-response predictions (Im omega = 0.042, Re omega = 0.846). The mode dislodges the central cusp, which subsequently settles on a slowly precessing elliptical orbit, and the l=1 power saturates into a long-lived oscillating 'soliton'. The authors interpret the growth and saturation as a gravitational analogue of the bump-on-tail instability, with resonant particles eroding the bump in the distribution function, and discuss implications for lopsidedness, off-center nuclei, and dynamical buoyancy.

Significance. If the central mechanism is established, the paper identifies a broad, previously underappreciated class of unstable equilibria within a standard family of density profiles and connects it to observable lopsidedness and off-center nuclei. The paper's strengths are credible: the dispersion-relation calculation is a parameter-free evaluation from the same distribution function used to seed the simulations, and the growth rate and pattern speed agree with the simulation at the few-percent level. The resolution and convergence tests in Appendix A, the use of multimass initial conditions with no mass segregation, and the direct orbital-frequency analysis (NAFF) are careful and add support to the linear-instability claim. The main weakness is that the bump-erosion saturation mechanism rests on an f(E,L) reconstruction that assumes spherical symmetry in a deliberately non-spherical saturated state; this part of the evidence needs strengthening or qualification.

major comments (3)
  1. [Section 3.3 and Figure 6] The computation of the evolving distribution function assumes spherical symmetry: g(E,L) in Eq. (11) is evaluated in the instantaneous spherically averaged potential, and f(E) is obtained by marginalizing over L via Eq. (13). The justification given in Section 3.3, namely that the l=1 mode saturates at a normalized amplitude of about 10^-2, uses the global A1/A0 amplitude, but the saturated state is locally strongly non-spherical: the cusp is displaced to r approximately 0.15 (Figure 5) and the central l=1 density contrast in Figure 4 is much larger than the global A1/A0. For particles librating in or near the displaced cusp, the pair (E,L) computed in a spherical potential is not a canonical map of the true phase space, so the apparent flattening of the bump in Figure 6 could be partly an artifact of the analysis. Because the plateau-formation/bump-erosion mechanism is the stated saturation mechanism in Section 6.2, the paper should either recompute the DF in action space using the full non-spherical potential or quantify the bias in the spherical (E,L) reconstruction.
  2. [Section 4.3 and Section 6.2] The manuscript reports that the late-time f(E) still has a small region with df/dE > 0 and attributes this to unmodeled anisotropy. This is in tension with the claim that saturation occurs when the bump has been eroded to df/dE roughly equal to zero: if a positive gradient remains, the linear-response growth mechanism would not be fully switched off, and the plateau-formation picture is incomplete. Moreover, the residual gradient is measured with the same spherically biased estimator as in Figure 6, so it is not clear whether it is real or an artifact. The authors should quantify the residual df/dE; if it is real, they should explain how saturation is achieved despite it, and if it is an artifact, the erosion conclusion needs to be revised accordingly.
  3. [Section 6.2.1 and Figure 11] The blanket statement in Section 2.1 that all simulated systems in the unstable region are found to be unstable is not supported by the alpha=4 run, which shows no detectable l=1 mode in Figure 11. The explanation that the mode amplitude is too small to rise above Poisson noise is plausible but not demonstrated: no predicted saturation amplitude or noise floor is given for alpha=4, and the resolution tests in Appendix A are for the fiducial (6,5,0.5) model only. The paper should either provide a quantitative estimate (linear-theory amplitude versus expected noise) for the alpha=4 case or restrict the general claim to systems with deeper inflections.
minor comments (4)
  1. [Section 4.2 and Figure 3 caption] There is a typo in Section 4.2: 'we use can the trajectory' should read 'we can use the trajectory'; also, the Figure 3 caption has 'a dipole perturpation', which should be 'a dipole perturbation'.
  2. [Section 6.4] The phrase 'as eluded to above' should be 'as alluded to above'.
  3. [Section 3.1, Equation (7)] The mass-spectrum exponent zeta is introduced as a free parameter, but the main runs use only zeta=15 and the Appendix compares equal-mass runs rather than varying zeta; a sentence on the sensitivity of the results to zeta would be helpful.
  4. [Figure 6] The left panel of Figure 6 would be easier to read if the color scale were saturated at a stated value of |Delta f| and if the Delta f = 0 contour were overlaid, since the sign of the change is central to the bump-erosion argument.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the linear-theory instability, growth rate, and pattern speed are independent, parameter-free computations from the same initial DF that seeds the simulations; self-citations are contextual, not load-bearing.

full rationale

The central derivation chain is: (i) choose a double power-law density profile; (ii) compute f(E) by Eddington inversion (Eq. 4); (iii) compute the linear response matrix M(omega) from that f(E) and locate the zero of D(omega) (Eqs. 1-2); (iv) sample N-body initial conditions from the same f(E) and evolve them. Steps (ii)-(iii) share the same input DF, but that is not circular: the response calculation could in principle have returned a damped mode, yet it yields a positive imaginary frequency, and the simulation independently measures growth rate 0.044 versus the predicted 0.042 and pattern speed 0.829 versus Re(omega) = 0.846. The saturation claim that the bump in the DF is eroded is an observational inference from the simulation, not a quantity fitted to the simulation and then called a prediction. The spherical reconstruction of f(E,L) in Section 3.3 is an approximation whose limitations are acknowledged in Section 4.3, including the residual positive df/dE and possible anisotropy; this is a correctness or robustness concern, not a circular reduction. Self-citations to Weinberg (1994, 2023) provide the numerical implementation of standard linear response theory and prior context, but the present conclusion does not rest on an unverified self-citation chain: the dispersion relation is evaluated explicitly by marching squares and compared against an independent N-body integration (GyrfalcOn) and a separate basis-function expansion (EXP). The fitted gamma_stable and gamma_physical lines are auxiliary classifiers of parameter space, not inputs to the instability calculation, so no fitted parameter is renamed as a prediction. No load-bearing step reduces, by the paper's own equations or by self-citation, to its own inputs.

Assumptions & free parameters 3 free parameters · 6 assumptions · 0 invented entities

The central claim depends on standard stellar-dynamical axioms (Eddington inversion, Antonov stability condition, collisionless Boltzmann, linear response theory) and on the modeling assumption that the simulated systems are exactly isotropic and remain spherical for the DF analysis. The fitted quantities in equations (5) and (6) and the mass-spectrum exponent zeta are auxiliary and do not enter the mode prediction.

free parameters (3)
  • gamma_stable fit coefficients = 0.0550, 0.087, 0.136, 0.181
    Coefficients in eq. (5) fit to the numerically computed stability boundary in (alpha,beta,gamma) space. Not used in the central instability demonstration, only to map the parameter space.
  • gamma_physical fit coefficients = 0.0047, 0.043, 0.013, 0.102
    Coefficients in eq. (6) fit to the physicality boundary where f(E) becomes negative. Secondary to the main claim.
  • mass spectrum exponent zeta = 15
    Equation (7) uses zeta=15 to increase central mass resolution. Verified not to affect results (Appendix A).
assumptions (6)
  • standard math Antonov's stability criterion: isotropic spherical systems with df/dE < 0 everywhere are stable.
    Used to define the 'unstable' region where the criterion is violated (Section 2.1).
  • standard math Eddington inversion determines f(E) uniquely from a spherically symmetric density profile.
    Equation (4) is used to compute the initial distribution function for each (alpha,beta,gamma) model.
  • domain assumption The collisionless Boltzmann equation and Poisson equation govern the system; N-body simulations with GyrfalcOn approximate their solutions.
    Standard modeling assumption for dark matter halos and galaxies.
  • domain assumption The system is isotropic, f = f(E), throughout the initial conditions.
    The paper restricts to isotropic models; this excludes anisotropies which could alter stability (stated in Section 7 as a limitation).
  • domain assumption The system remains spherical during the evolution, so the density of states can be computed from the spherically averaged potential.
    Section 3.3; justified by the mode amplitude being less than about 1 percent, but this is a modeling assumption for inferring f(E,L).
  • standard math The linear response matrix formalism (Kalnajs, Weinberg) correctly predicts discrete modes of the self-gravitating system.
    Section 4.4 uses the response matrix (eq. 1) to find the l=1 mode frequency and growth rate.

how reviews work

0 comments
Cite this review

Pith. "Pith review of The Dipole Instability in Gravitational $N$-body Systems: A Natural Explanation for Lopsidedness and Off-Centered Nuclei in Galaxies." pith.science (2026). https://pith.science/paper/EIRXXFLV

@misc{pith2026250523905,
  author       = {Pith},
  title        = {Pith review of: The Dipole Instability in Gravitational $N$-body Systems: A Natural Explanation for Lopsidedness and Off-Centered Nuclei in Galaxies},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/EIRXXFLV}},
  note         = {Machine review of arXiv:2505.23905}
}
abstract

We explore the stability of isotropic, spherical, self-gravitating systems with a double-power law density profile. Systems with rapid transitions between the inner and outer slopes are shown to have an inflection in their isotropic distribution function (DF), where ${\rm d} f/{\rm d} E > 0$, thereby violating Antonov's stability criterion. Using high-resolution $N$-body simulations, we show that the resulting instability causes the growth of a rotating dipole (or $l=1$) mode. The inflection feature in the DF responds to the mode by promoting its growth, driving the instability. The growth of the dipole results in a torque that dislodges the original cusp from its central location, and sets it in motion throughout the central region. Once the mode goes non-linear, it saturates, together with the cusp, into a long-lived soliton (the $l=1$ equivalent of a bar in a disk galaxy), which maintains its sloshing motion through the center of the halo along a slowly precessing, elliptical orbit. Concurrently, the soliton traps increasingly more particles into libration, and the exchange of energy and angular momentum with these trapped particles works towards eroding the bump in the distribution function. We point out similarities between the dipole mode and the bump-on-tail instability in electrostatic plasmas, and highlight a potential connection with core stalling and dynamical buoyancy in systems with a cored density profile. Finally, we discuss the astrophysical implications in terms of lopsidedness and off-center nuclei in galaxies.

Figures

Figures reproduced from arXiv: 2505.23905 by the authors.

Figure 1
Figure 1. [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. These figures show the regions in 𝛼 − 𝛾 parameter space, for fixed 𝛽, where the system is stable (d 𝑓 /d𝐸 < 0 for all 𝐸, shaded yellow), unstable (d 𝑓 /d𝐸 > 0 at some 𝐸 but with 𝑓 (𝐸) ≥ 0 at all 𝐸, shaded green), and unphysical ( 𝑓 (𝐸) < 0 at some 𝐸, shaded purple). Different panels correspond to three different values of 𝛽 as indicated. For comparison, some common density profiles are indicated, including the NFW p… view at source ↗
Figure 3
Figure 3. Projected over/underdensity (ΔΣ = Σ − Σ0) of our fiducial simulation in the 𝑥-𝑦 plane. Different panels correspond to snapshots at different time 𝑇, as indicated. At 𝑇 = 70 the signatures a dipole perturpation begin to emerge, which is clearly visible by 𝑇 = 100. By 𝑇 = 150 the 𝑙 = 1 mode has saturated (see [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (9 more)
Figure 4
Figure 4. Figure 4: Left panel: the temporal and spatial evolution of the 𝑙 = 1 harmonic mode, ⟨𝐶1 (𝑟)⟩, computed using equation (17) and normalized by the 𝑙 = 0 amplitude ⟨𝐶0 (𝑟)⟩. Time is on the 𝑥-axis and radius from the center of mass, 𝑟, is along the 𝑦-axis, while the color coding re…
Figure 5
Figure 5. Figure 5: Top panel: the trajectory of the cusp in the 𝑥-𝑦 plane color coded by time, with the origin fixed at the center of mass and the z-axis pointing along the cusp’s angular momentum vector. Bottom panel: radial distance of the cusp from the center of mass of the system as …
Figure 6
Figure 6. Figure 6: Left panel: the change in the two-dimensional DF, Δ 𝑓 (𝐸, ℓ), between 𝑇 = 500 and 𝑇 = 0. Right panel: the evolution of 𝑓 (𝐸) (equation [13]) over time. Note how the bump in 𝑓 (𝐸) flattens out over time as the 𝑙 = 1 mode grows. As is evident from the left panel, this er…
Figure 7
Figure 7. Figure 7: Contours of the dispersion relation |D (𝜔)| of our fidu￾cial system for the 𝑙 = 1 harmonic in the complex 𝜔 plane. The zero of the |D (𝜔)| occurs at 𝜔mode = (0.846, 0.042), indicated by the white cross. For comparison, the horizontal green dashed line corresponds to Im…
Figure 8
Figure 8. Figure 8: Top panel: frequency diagram plotting Ω𝜙 vs. Ω𝑟 for a random set of 105 particles in the simulation. Each point is color coded by the maximum range in energy, (Δ𝐸)max = 𝐸max − 𝐸min, experienced by the particle over the entire duration of the simulation. Resonances are …
Figure 9
Figure 9. Figure 9: An example of a near-resonant orbit in the simulation. Left panel: trajectory in the inertial frame. Middle panel: trajectory in the frame co-rotating with the cusp. The red bar indicates the extent of the radial oscillations of the cusp (cf [PITH_FULL_IMAGE:figures/f…
Figure 10
Figure 10. Figure 10: Map of the average fractional change in energy, Δ𝐸 = 𝐸(𝑇) − 𝐸0 (color coded), as a function of time, 𝑇, and initial energy 𝐸0. The average is computed at each simulation output by averaging Δ𝐸 over all particles in small bins of 𝐸0. Around 𝑇 ≈ 100 when the 𝑙 = 1 start…
Figure 11
Figure 11. Figure 11: The normalized 𝑙 = 1 amplitude of the gravitational power, 𝐴1/𝐴0, computed with EXP as a function of time. Results are shown for simulations of (𝛼, 𝛽, 𝛾) systems with 𝛽 = 5.0 and 𝛾 = 0.5, but different values of 𝛼, as indicated. The inset shows the corresponding initi…
Figure 12
Figure 12. Figure 12: Left panel: the average mass of simulation particles as a function of distance from the center of mass at 𝑇 = 0 (orange) and 𝑇 = 500 (black dotted) for a simulation of a spherical system with an initial Hernquist density profile. The two lines lie exactly on top of ea…

Discussion (0). Sign in to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Constraining the Milky Way dark matter halo with LMC-induced reflex motion

    astro-ph.GA 2025-08 unverdicted novelty 5.0 of 10

    Basis-function-expansion N-body models show that the direction, not the amplitude, of the LMC-induced reflex motion of the Milky Way disc can constrain the outer dark matter halo, and reveal a slope-dependent halo ins...

Reference graph

Works this paper leans on

86 extracted references · 24 canonical work pages · cited by 1 Pith paper

  1. [1]

    Antonov, V. A. 1962, Solution of the problem of stability of stellar system Emden’s density law and the spherical distribution of velocities

    1962

  2. [2]

    Antonov, V. A. 1973, in Dynamics of Galaxies and Star Clusters, 139–143

    1973

  3. [3]

    2021, MNRAS, 503, 2955, doi: 10.1093/mnras/stab634

    Baes, M., & Camps, P. 2021, MNRAS, 503, 2955, doi: 10.1093/mnras/stab634

    work page doi:10.1093/mnras/stab634 2021

  4. [4]

    Banik, U., & van den Bosch, F. C. 2021, ApJ, 912, 43, doi: 10.3847/1538-4357/abeb6d —. 2022, ApJ, 926, 215, doi: 10.3847/1538-4357/ac4242

    doi:10.3847/1538-4357/abeb6d 2021

  5. [5]

    1986, ApJ, 300, 112, doi: 10.1086/163786

    Barnes, J., Goodman, J., & Hut, P. 1986, ApJ, 300, 112, doi: 10.1086/163786

    work page doi:10.1086/163786 1986

  6. [6]

    C., Blandford, R

    Begelman, M. C., Blandford, R. D., & Rees, M. J. 1980, Nature, 287, 307, doi: 10.1038/287307a0 Beraldo e Silva, L., Debattista, V. P., Anderson, S. R., et al. 2023, ApJ, 955, 38, doi: 10.3847/1538-4357/ace976

    doi:10.1038/287307a0 1980

  7. [7]

    2000, A&A, 359, 447

    Binggeli, B., Barazza, F., & Jerjen, H. 2000, A&A, 359, 447

    2000

  8. [8]

    1988, MNRAS, 230, 597, doi: 10.1093/mnras/230.4.597

    Binney, J., & Lacey, C. 1988, MNRAS, 230, 597, doi: 10.1093/mnras/230.4.597

    doi:10.1093/mnras/230.4.597 1988

Show all 86 references
  1. [9]

    2008, Galactic Dynamics: Second Edition

    Binney, J., & Tremaine, S. 2008, Galactic Dynamics: Second Edition

    2008

  2. [10]

    J., & Puerari, I

    Bournaud, F., Combes, F., Jog, C. J., & Puerari, I. 2005, A&A, 438, 507, doi: 10.1051/0004-6361:20052631

    2005 doi

  3. [11]

    Boyd, T. J. M., & Sanderson, J. J. 2003, The Physics of Plasmas, doi: 10.1017/CBO9780511755750

    2003 doi

  4. [12]

    1943, ApJ, 97, 255, doi: 10.1086/144517

    Chandrasekhar, S. 1943, ApJ, 97, 255, doi: 10.1086/144517

    1943 doi

  5. [13]

    2016, A&A, 588, A48, doi: 10.1051/0004-6361/201527730

    Chemin, L., Hur´e, J.-M., Soubiran, C., et al. 2016, A&A, 588, A48, doi: 10.1051/0004-6361/201527730

    2016 doi

  6. [14]

    T., van den Bosch, F

    Chiang, B. T., van den Bosch, F. C., & Schive, H.-Y. 2024, arXiv e-prints, arXiv:2411.03192, doi: 10.48550/arXiv.2411.03192

    2024 doi

  7. [15]

    2023, MNRAS, 525, 3576, doi: 10.1093/mnras/stad2324

    Chiba, R. 2023, MNRAS, 525, 3576, doi: 10.1093/mnras/stad2324

    2023 doi

  8. [16]

    2022, MNRAS, 513, 768, doi: 10.1093/mnras/stac697

    Chiba, R., & Sch¨onrich, R. 2022, MNRAS, 513, 768, doi: 10.1093/mnras/stac697

    2022 doi

  9. [18]

    D., van Dokkum, P., et al

    Conroy, C., Johnson, B. D., van Dokkum, P., et al. 2024, ApJ, 968, 129, doi: 10.3847/1538-4357/ad414a

    2024 doi

  10. [19]

    J., & Wyse, R

    Daniel, K. J., & Wyse, R. F. G. 2015, MNRAS, 447, 3576, doi: 10.1093/mnras/stu2683

    2015 doi

  11. [20]

    M., & Shanny, R

    Dawson, J. M., & Shanny, R. 1968, Physics of Fluids, 11, 1506, doi: 10.1063/1.1692136 de Zeeuw, T. 1985, MNRAS, 216, 273, doi: 10.1093/mnras/216.2.273

    1968 doi

  12. [21]

    1993, MNRAS, 265, 250, doi: 10.1093/mnras/265.1.250 —

    Dehnen, W. 1993, MNRAS, 265, 250, doi: 10.1093/mnras/265.1.250 —. 2001, MNRAS, 324, 273, doi: 10.1046/j.1365-8711.2001.04237.x —. 2002, Journal of Computational Physics, 179, 27, doi: 10.1006/jcph.2002.7026

    1993

  13. [22]

    P., & Feix, M

    Doremus, J. P., & Feix, M. R. 1972, A&A, 20, 259

    1972

  14. [23]

    R., & Baumann, G

    Doremus, J.-P., Feix, M. R., & Baumann, G. 1971, Phys. Rev. Lett., 26, 725, doi: 10.1103/PhysRevLett.26.725

    1971 doi

  15. [24]

    E., Taylor, J

    Drakos, N. E., Taylor, J. E., & Benson, A. J. 2017, MNRAS, 468, 2345, doi: 10.1093/mnras/stx652 —. 2020, MNRAS, 494, 378, doi: 10.1093/mnras/staa760

    2017 doi

  16. [26]

    Eddington, A. S. 1916, MNRAS, 76, 572, doi: 10.1093/mnras/76.7.572

    1916 doi

  17. [27]

    2001, ApJ, 560, 636, doi: 10.1086/322516

    El-Zant, A., Shlosman, I., & Hoffman, Y. 2001, ApJ, 560, 636, doi: 10.1086/322516

    2001 doi

  18. [28]

    B., Pichon, C., & Chavanis, P

    Fouvry, J. B., Pichon, C., & Chavanis, P. H. 2015, A&A, 581, A139, doi: 10.1051/0004-6361/201525928

    2015 doi

  19. [29]

    M., & Polyachenko, V

    Fridman, A. M., & Polyachenko, V. L. 1984, Physics of gravitating systems. II - Nonlinear collective processes: Nonlinear waves, solitons, collisionless shocks, turbulence. Astrophysical applications

    1984

  20. [30]

    I., & Stadel, J

    Goerdt, T., Moore, B., Read, J. I., & Stadel, J. 2010, ApJ, 725, 1707, doi: 10.1088/0004-637X/725/2/1707

    2010 doi

  21. [32]

    1971, Ap&SS, 13, 387, doi: 10.1007/BF00649168

    Goldstein, S. 1971, Ap&SS, 13, 387, doi: 10.1007/BF00649168

    1971 doi

  22. [33]

    2024, MNRAS, doi: 10.1093/mnras/stae305

    Hamilton, C. 2024, MNRAS, doi: 10.1093/mnras/stae305

    2024 doi

  23. [34]

    2024, arXiv e-prints, arXiv:2402.13322, doi: 10.48550/arXiv.2402.13322

    Hamilton, C., & Fouvry, J.-B. 2024, arXiv e-prints, arXiv:2402.13322, doi: 10.48550/arXiv.2402.13322

    2024 doi arXiv

  24. [35]

    A., Arzamasskiy, L., & Duarte, V

    Hamilton, C., Tolman, E. A., Arzamasskiy, L., & Duarte, V. N. 2023, ApJ, 954, 12, doi: 10.3847/1538-4357/acd69b

    2023 doi

  25. [36]

    P., Hogg, D

    Haynes, M. P., Hogg, D. E., Maddalena, R. J., Roberts, M. S., & van Zee, L. 1998, AJ, 115, 62, doi: 10.1086/300166

    1998 doi

  26. [37]

    C., Breen, P

    Heggie, D. C., Breen, P. G., & Varri, A. L. 2020, MNRAS, 492, 6019, doi: 10.1093/mnras/staa272

    2020 doi

  27. [38]

    1990, ApJ, 356, 359, doi: 10.1086/168845

    Hernquist, L. 1990, ApJ, 356, 359, doi: 10.1086/168845

    1990 doi

  28. [40]

    1983, MNRAS, 202, 995, doi: 10.1093/mnras/202.4.995

    Jaffe, W. 1983, MNRAS, 202, 995, doi: 10.1093/mnras/202.4.995

    1983 doi

  29. [41]

    Jog, C. J. 1997, ApJ, 488, 642, doi: 10.1086/304721 —. 1999, ApJ, 522, 661, doi: 10.1086/307664

    1997 doi

  30. [42]

    J., & Combes, F

    Jog, C. J., & Combes, F. 2009, PhR, 471, 75, doi: 10.1016/j.physrep.2008.12.002

    2009 doi

  31. [43]

    Kalnajs, A. J. 1977a, ApJ, 212, 637, doi: 10.1086/155086 —. 1977b, ApJ, 212, 637, doi: 10.1086/155086

    doi

  32. [44]

    2018, ApJ, 868, 134, doi: 10.3847/1538-4357/aaeacf

    Kaur, K., & Sridhar, S. 2018, ApJ, 868, 134, doi: 10.3847/1538-4357/aaeacf

    2018 doi

  33. [45]

    King, I. R. 1966, AJ, 71, 64, doi: 10.1086/109857

    1966 doi

  34. [46]

    2013, ApJ, 765, 10, doi: 10.1088/0004-637X/765/1/10

    Kuhlen, M., Guedes, J., Pillepich, A., Madau, P., & Mayer, L. 2013, ApJ, 765, 10, doi: 10.1088/0004-637X/765/1/10

    2013 doi

  35. [47]

    1992, Physica D Nonlinear Phenomena, 56, 253, doi: 10.1016/0167-2789(92)90028-L

    Laskar, J., Froeschl´e, C., & Celletti, A. 1992, Physica D Nonlinear Phenomena, 56, 253, doi: 10.1016/0167-2789(92)90028-L

    1992 doi

  36. [48]

    E., & Sparke, L

    Levine, S. E., & Sparke, L. S. 1998, ApJL, 496, L13, doi: 10.1086/311239

    1998 doi

  37. [49]

    Lynden-Bell, D., & Kalnajs, A. J. 1972, MNRAS, 157, 1, doi: 10.1093/mnras/157.1.1

    1972 doi

  38. [50]

    2015, MNRAS, 449, 3393, doi: 10.1093/mnras/stv467

    Massey, R., Williams, L., Smit, R., et al. 2015, MNRAS, 449, 3393, doi: 10.1093/mnras/stv467

    2015 doi

  39. [51]

    McMillan, P. J. 2017, MNRAS, 465, 76, doi: 10.1093/mnras/stw2759

    2017 doi

  40. [52]

    F., Santos-Santos, I., Ben´ıtez-Llambay, A., & Frenk, C

    Meadows, N., Navarro, J. F., Santos-Santos, I., Ben´ıtez-Llambay, A., & Frenk, C. 2020, MNRAS, 491, 3336, doi: 10.1093/mnras/stz3280

    2020 doi

  41. [53]

    Merritt, D., & Aguilar, L. A. 1985, MNRAS, 217, 787, doi: 10.1093/mnras/217.4.787

    1985 doi

  42. [54]

    2024, MNRAS, 528, 5252, doi: 10.1093/mnras/stae292

    Mezcua, M., & Dom´ınguez S´anchez, H. 2024, MNRAS, 528, 5252, doi: 10.1093/mnras/stae292

    2024 doi

  43. [55]

    Michie, R. W. 1963, MNRAS, 125, 127, doi: 10.1093/mnras/125.2.127

    1963 doi

  44. [56]

    H., & Smith, B

    Miller, R. H., & Smith, B. F. 1992, ApJ, 393, 508, doi: 10.1086/171523

    1992 doi

  45. [57]

    T., Nuritdinov, S

    Mirtadjieva, K. T., Nuritdinov, S. N., Ruzibaev, J. K., & Khalid, M. 2011, Astrophysics, 54, 184, doi: 10.1007/s10511-011-9171-z

    2011 doi

  46. [58]

    F., Frenk, C

    Navarro, J. F., Frenk, C. S., & White, S. D. M. 1997, ApJ, 490, 493, doi: 10.1086/304888

    1997 doi

  47. [60]

    L., & Papaloizou, J

    Palmer, P. L., & Papaloizou, J. 1987, MNRAS, 224, 1043, doi: 10.1093/mnras/224.4.1043

    1987 doi

  48. [61]

    1996, A&A, 307, 427

    Papaphilippou, Y., & Laskar, J. 1996, A&A, 307, 427

    1996

  49. [62]

    S., Roule, M., Fouvry, J.-B., Pichon, C., & Tep, K

    Petersen, M. S., Roule, M., Fouvry, J.-B., Pichon, C., & Tep, K. 2024, MNRAS, 530, 4378, doi: 10.1093/mnras/stae732

    2024 doi

  50. [63]

    S., & Weinberg, M

    Petersen, M. S., & Weinberg, M. D. 2025, Journal of Open Source Software, 10, 7302, doi: 10.21105/joss.07302

    2025 doi

  51. [64]

    S., Weinberg, M

    Petersen, M. S., Weinberg, M. D., & Katz, N. 2022, MNRAS, 510, 6201, doi: 10.1093/mnras/stab3639

    2022 doi

  52. [65]

    A., Gualandris, A., & Read, J

    Petts, J. A., Gualandris, A., & Read, J. I. 2015, MNRAS, 454, 3778, doi: 10.1093/mnras/stv2235

    2015 doi

  53. [66]

    A., Read, J

    Petts, J. A., Read, J. I., & Gualandris, A. 2016, MNRAS, 463, 858, doi: 10.1093/mnras/stw2011

    2016 doi

  54. [67]

    Plummer, H. C. 1911, MNRAS, 71, 460, doi: 10.1093/mnras/71.5.460

    1911 doi

  55. [68]

    L., & Shukhman, I

    Polyachenko, V. L., & Shukhman, I. G. 1981, Soviet Ast., 25, 533

    1981

  56. [69]

    2012, MNRAS, 421, 3464, doi: 10.1111/j.1365-2966.2012.20571.x

    Pontzen, A., & Governato, F. 2012, MNRAS, 421, 3464, doi: 10.1111/j.1365-2966.2012.20571.x

    2012

  57. [70]

    D., & Hernquist, L

    Quinlan, G. D., & Hernquist, L. 1997, NewA, 2, 533, doi: 10.1016/S1384-1076(97)00039-0

    1997 doi

  58. [71]

    I., Goerdt, T., Moore, B., et al

    Read, J. I., Goerdt, T., Moore, B., et al. 2006, MNRAS, 373, 1451, doi: 10.1111/j.1365-2966.2006.11022.x

    2006

  59. [72]

    2008, ApJ, 677, 186, doi: 10.1086/526506

    Kauffmann, G. 2008, ApJ, 677, 186, doi: 10.1086/526506

    2008 doi

  60. [73]

    E., Condon, J

    Reines, A. E., Condon, J. J., Darling, J., & Greene, J. E. 2020, ApJ, 888, 36, doi: 10.3847/1538-4357/ab4999

    2020 doi

  61. [74]

    G., & Sancisi, R

    Richter, O. G., & Sancisi, R. 1994, A&A, 290, L9

    1994

  62. [75]

    S., Jog, C

    Saha, K., Levine, E. S., Jog, C. J., & Blitz, L. 2009, ApJ, 697, 2015, doi: 10.1088/0004-637X/697/2/2015

    2009 doi

  63. [76]

    Sellwood, J. A. 2008, ApJ, 679, 379, doi: 10.1086/586882 22 —. 2014, Reviews of Modern Physics, 86, 1, doi: 10.1103/RevModPhys.86.1

    2008 doi

  64. [77]

    A., & Binney, J

    Sellwood, J. A., & Binney, J. J. 2002, MNRAS, 336, 785, doi: 10.1046/j.1365-8711.2002.05806.x

    2002

  65. [78]

    A., & Wilkinson, A

    Sellwood, J. A., & Wilkinson, A. 1993, Reports on Progress in Physics, 56, 173, doi: 10.1088/0034-4885/56/2/001

    1993 doi

  66. [79]

    Sigurdsson, S., Hernquist, L., & Quinlan, G. D. 1995, ApJ, 446, 75, doi: 10.1086/175769

    1995 doi

  67. [80]

    Albada, T. S. 1999, MNRAS, 304, 330, doi: 10.1046/j.1365-8711.1999.02332.x

    1999

  68. [81]

    1995, in Astronomical Society of the Pacific Conference

    Teuben, P. 1995, in Astronomical Society of the Pacific Conference

    1995

  69. [82]

    Teyssier, R., Pontzen, A., Dubois, Y., & Read, J. I. 2013, MNRAS, 429, 3068, doi: 10.1093/mnras/sts563

    2013 doi

  70. [83]

    O., Byun, Y.-I., et al

    Tremaine, S., Richstone, D. O., Byun, Y.-I., et al. 1994, AJ, 107, 634, doi: 10.1086/116883

    1994 doi

  71. [84]

    Tremaine, S., & Weinberg, M. D. 1984, MNRAS, 209, 729, doi: 10.1093/mnras/209.4.729

    1984 doi

  72. [85]

    P., Quinn, T., & Moore, B

    Valluri, M., Debattista, V. P., Quinn, T., & Moore, B. 2010, MNRAS, 403, 525, doi: 10.1111/j.1365-2966.2009.16192.x

    2010

  73. [86]

    P., Quinn, T

    Valluri, M., Debattista, V. P., Quinn, T. R., Roˇskar, R., & Wadsley, J. 2012, MNRAS, 419, 1951, doi: 10.1111/j.1365-2966.2011.19853.x

    2012

  74. [87]

    1998, ApJ, 506, 686, doi: 10.1086/306269

    Valluri, M., & Merritt, D. 1998, ApJ, 506, 686, doi: 10.1086/306269

    1998 doi

  75. [88]

    Weinberg, M. D. 1989, MNRAS, 239, 549, doi: 10.1093/mnras/239.2.549 —. 1991, ApJ, 368, 66, doi: 10.1086/169671 —. 1994, ApJ, 421, 481, doi: 10.1086/173665 —. 2004, arXiv e-prints, astro, doi: 10.48550/arXiv.astro-ph/0404169 —. 2023, MNRAS, 525, 4962, doi: 10.1093/mnras/stad2591

    1989 doi arXiv

  76. [89]

    2013, ApJ, 772, 135, doi: 10.1088/0004-637X/772/2/135

    Zaritsky, D., Salo, H., Laurikainen, E., et al. 2013, ApJ, 772, 135, doi: 10.1088/0004-637X/772/2/135

    2013 doi

  77. [91]

    2008, MNRAS, 387, 1719, doi: 10.1111/j.1365-2966.2008.13362.x

    Zhang, M., & Magorrian, J. 2008, MNRAS, 387, 1719, doi: 10.1111/j.1365-2966.2008.13362.x

    2008

  78. [92]

    baseline

    Zhao, H. 1996, MNRAS, 278, 488, doi: 10.1093/mnras/278.2.488 23 Figure 12.Left panel: the average mass of simulation particles as a function of distance from the center of mass at𝑇=0 (orange) and𝑇=500 (black dotted) for a simulation of a spherical system with an initial Hernqu...

    1996 doi

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.